Thus do they all: a model for R&D investment

Our study introduces an under-reported determinant of Research and Development investments: interaction among companies. We measure the interaction using a statistical mechanics model on data from U.S. pharmaceutical and biotechnology companies (1989-2012). Our results show that firms tend to influence each other when determining their level of R&D investment. This means that there is a variable affecting the investment behavior which recalls the concept of herding. Our model is descriptive of the status quo of the R&D investment behavior and it also allows predictions to be made at industry and firm level. In fact, it measures how a variation of the interaction with other related firms, and the exogenous conditions of their economic environment can affect a firm's investment choice.